A central problem in biochemistry is the relationship between amino acid sequence and the native three-dimensional structure adopted by a peptide or a protein in a particular environment. The proposed research addresses this problem through the complementary use of computer analysis of known protein structures, and design, synthesis, and conformational analysis of model peptides. The sequences of the model peptides will be chosen based on results of a computer-assisted search for structural features of interest in native proteins, and based on previous work in this laboratory in which peptide models of well-defined conformations were developed. The proposed studies will include nonrepetitive structural features, such as reverse turns, that are regular, identifiable local conformations, but are not part of periodic structures (e.g. Alpha-helix or Beta-structure), and hence cannot be described by Phi,Psi angles that repeat along the polypeptide chain. It is further proposes to study the conformational impact of three environments that occur in and near a biological membrane: an interfacial region between an aqueous phase and a hydrophobic phase, the unique water adjacent to a charged interface, and a hydrophobic microenvironment. Conformational studies will be done on the peptide models solubilized in normal detergent micelles, in reversed micelles, and in vesicles. Sequences likely to reside near an interface or within a hydrophobic phase will be identified by searching the surface and the interior regions of globular proteins of known structure, and designing model peptides from the sequences that are observed to take up structural features of interest. Results of the proposed studies will be used 1) to predict from amino acid sequences the sites of structural features in membrane proteins, and where within the membrane they would tend to reside, and 2) to predict, in a more sophisticated way than is presently possible, surface and interior regions and local conformations of globular proteins. Such predictive capability makes possible the identification of likely antigenic determinants, recognition sites, hydrophobic anchors, and other functionally important structural features in proteins of known sequence.